270 likes | 401 Views
Luminosity Monitoring and Beam Diagnostics. Christian Grah. FCAL Collaboration Workshop TAU, September 18-19, 2005. Outline. Very Forward Calorimetry of LDC, LumiCal PhotoCal BeamCal Fast luminosity monitoring Beam diagnostics from beamstrahlung analyzing pairs and photons
E N D
Luminosity MonitoringandBeam Diagnostics Christian Grah FCAL Collaboration Workshop TAU, September 18-19, 2005
Outline • Very Forward Calorimetry of LDC, • LumiCal • PhotoCal • BeamCal • Fast luminosity monitoring • Beam diagnostics from beamstrahlung • analyzing pairs and photons • 20 mrad crossing angle • Summary & outlook Ch.Grah: Beamdiagnostics
Very Forward Region LumiCal: 26 < θ < 82 mrad BeamCal: 4 < q < 28 mrad PhotoCal: 100 < q < 400 μrad Ch.Grah: Beamdiagnostics
Very Forward Calorimeters • LumiCal: • Precise measurement of the luminosity by using bhabha events (very high mechanical precision needed). • Extend coverage of the detector system. • Photocal • beam diagnostics from beamstrahlung photons • BeamCal: • detection of electrons/photons at low angle. • shielding of Inner Detector. • beam diagnostics from beamstrahlung electrons/positron pairs. Ch.Grah: Beamdiagnostics
BeamCal: Beam Diagnosticsand Fast Luminosity Monitoring e+ e- e+e-pairs from beamstrahlung are deflected into the BeamCal Deposited energy from pairs at z = +365 (no B-field, TESLA parameters) • 15000 e+e- per BX => 10 – 20 TeV • ~ 10 MGy per year • “fast” => O(μs) • Direct photons for q< 400 mrad (PhotoCal) Ch.Grah: Beamdiagnostics
Distribution of pairs Bz=4T Bz=4T/2mrad Ch.Grah: Beamdiagnostics
BeamCal: W-Diamond Sandwich Space for electronics Length = 30 X0 (3.5mm W + .5mm diamond sensor) ~ 15 000 channels ~1.5/2 cm < R < ~10(+2) cm Ch.Grah: Beamdiagnostics
Fast Luminosity Monitoring A fast estimation of the luminosity is needed, e.g. number of pairs/total energy. • This will be included into the fast feedback system. Luminosity development during first 600 bunches of a bunch-train. Ltotal = L(1-600) + L(550600)*(2820-600)/50 G.White QMUL/SLAC RHUL & Snowmass presentation position and angle scan Ch.Grah: Beamdiagnostics
L Improvement by including Pair Signal G.White position scan position scan & angle scan 500 GeV 350 GeV Up to 14.8 % improvement on the overall luminosity. A fast luminosity signal is a must have for the ILC. Ch.Grah: Beamdiagnostics
Beamstrahlung Pairs • Observables (examples): • total energy • first radial moment • thrust value • angular spread • E(ring ≥ 4) / Etot • E / N • left/right, up/down, forward/backward asymmetries detector: realistic segmentation, ideal resolution bunch by bunch resolution Ch.Grah: Beamdiagnostics
Aim: Reconstruction of Beam Parameters Sigma x Sigma y Sigma z Sigma x' Sigma y' x offset y offset x' offset y' offset x-waist shift y-waist shift Bunch rotation N particles/bunch Banana shape … … Luminosity What do you want to know? Ch.Grah: Beamdiagnostics
Analysis Concept Taylor Matrix Observables Observables Δ BeamPar nom • Beam Parameters • determine collision • creation of beamstr. • creation of e+e- pairs • guinea-pig • (D.Schulte) • Observables • characterize energy distributions in detectors • FORTRAN • analysis program (A.Stahl) 1st order Taylor-Exp. Solve by matrix inversion (Moore-Penrose Inverse) = + * Ch.Grah: Beamdiagnostics
Slopes parametrization (polynomial) 1 point = 1 bunch crossing by guinea-pig slope at nom. value taylor coefficient i,j observable j [au] beam parameter i [au] Ch.Grah: Beamdiagnostics
Analysis Test with non-nominal bunches: e- e+ nom. bunch size x: 575nm 575nm 553nm bunch size y: 5nm 7nm 5nm bunch size z: 290μm 320μm 300μm Ch.Grah: Beamdiagnostics
Multi Parameter Analysis σx Δσx σy Δσy σz Δσz 0.3 % 0.4 % 3.4 % 9.5 % 1.4 % 0.8 % 1.5 % 0.9 % 0.3 % 0.4 % 3.5 % 11 % 0.9 % 1.0 % 11 % 24 % 5.7 % 24 % 1.6 % 1.9 % 1.8 % 1.1 % 16 % 27 % 3.2 % 2.1 % Ch.Grah: Beamdiagnostics
2nd Order Calculations Fast feedback system group (QMUL) is studying higher order dependencies. Computation of Taylor Matrix only to second order, due to cpu limitations: 2nd order requires 7 days on 100 CPU cluster. This effort aims for an offline analysis/fitting routine. G.White Ch.Grah: Beamdiagnostics
Photons from Beamstrahlung • Report on analysis done by Andrei Rybin (Protvino) • Beam parameters under investigation: • beam size σx, σy, σz • beam offset x,y • waist shift x,y • Observables • Eγ (left – right) along x • Eγ (up – down) along y • Eγ (forward – backward) along z Every beam parameter analyzed separately, all others were fixed to nominal value. Ch.Grah: Beamdiagnostics
Interlude: Photon Distribution and Selection Photon selection:|angle x| > 0.2 mrad|angle y| > 0.1 mrad Etot=3366.83 TeV In use: Multi layer perceptron feed-forward network. 200 Epochs for learning hybrid linear BFGS learning method Ch.Grah: Beamdiagnostics
Vertical Beam Waist Variation vert. waist shift > 0 vert. waist shift < 0 nominell beam parameters Ch.Grah: Beamdiagnostics
Results of Analysis Parameter nom. in phot. pairs • σx 553 nm 4.2 1.5 • σy 5.0 nm 0.1 0.2 • σz 300 μm 7.5 4.3 • beam offset x 0 nm 4.0 6.0 • beam offset y 0 nm 0.16 0.4 • vertical waist shift 360 μm 14 24 Ch.Grah: Beamdiagnostics
Moving to 20mrad crossing angle Simplification of DID field map (B.Parker & A.Seryi) Ch.Grah: Beamdiagnostics
20mrad crossing angle 20mrad crossing angle and serpentine B-field implemented. |B| = 4T, serpentine field, 20mrad Ch.Grah: Beamdiagnostics
Geometry for 20mrad mirrored Ch.Grah: Beamdiagnostics
Beamdiagnostics for 20mrad • Simulate collision/beamstrahlung with Guineapig using head-on case. • Use FORTRAN code by A.Stahl - extended by • Boosting the generated pairs according to crossing angle. • Calculate impact position at BeamCal front face by parameterized movement in constant b-field. • Use new geometries (headon, 2mrad, 20mrad). • Used old set of observables (space for improvements). Ch.Grah: Beamdiagnostics
Single Parameter Analysis for 20mrad Ch.Grah: Beamdiagnostics
Summary • A fast luminosity signal could significantly increase luminosity. • Analyzing beamstrahlung grants access to many beam parameters. • Investigating the possibility of measuring the distribution of pairs from beamstrahlung with BeamCal. Promising results also for 20mrad case, but further work needed for final result. • Single and Multiparameter analysis is feasible. • Photon distribution at lowest angles (~100 μrad) with PhotoCal has potential. The more uncorrelated observables the better our parameter reconstruction will work. Ch.Grah: Beamdiagnostics
Outlook • Optimization of observables. • Comparison of the simple b-field approximation to using the b-field map. • Include a realistic detector response. • Look at realistic beam simulation and non linear bunch effects. Ch.Grah: Beamdiagnostics